fix(openai): Fast 模式对 GPT 模型重新生效（Responses 路径）

CHANGELOG.md

@@ -9,6 +9,7 @@ This changelog tracks **Minis for Android** as an independently maintained Andro
 - `android-shizuku-cli` is now `android-root-cli`: the same subcommands and JSON, run as root through `su`. Its Agent permission switch carries over.
 - The virtual screen runs in a root service (libsu) instead of a Shizuku UserService. Run its compatibility check once more to switch it back on.
 - Accessibility repair, clearing restricted settings and one-tap authorize use Root; Settings > System has a Root page in place of "Root & Shizuku", and the readiness checklist no longer lists Shizuku.
+- Fast Mode works for GPT models again on the Responses path (Codex login and Responses relays): with the toggle on, requests carry `service_tier: "priority"`, as OpenMinis 1.14 does. A change that kept only the xAI case had turned the toggle into a no-op for GPT.
 
 ## 0.0.2 — 2026-10-10
 

src/android/app/src/main/java/com/openminis/app/provider/openai/OpenAIProviderResponses.kt

@@ -209,7 +209,19 @@ internal fun OpenAIProvider.buildResponsesAPIBody(
     // through, so no isOAuth narrowing. Ineligible upstreams ignore the
     // field or silently downgrade (receipt visible via the
     // response.completed service_tier log).
+    //
+    // [T-android-xai-priority] The per-instance (xAI) tier is resolved first and the
+    // gpt-family rule only fills a missing key, so neither clobbers the other. The two
+    // gates are disjoint today: that toggle is xAI-only and xAI never reaches this builder.
+    // The gpt-family rule had been replaced by the xAI call alone, which made Fast Mode a
+    // silent no-op for every GPT model (OpenMinis 1.14 keeps both).
     resolvedServiceTier()?.let { body.put("service_tier", it) }
+    if (!body.has("service_tier") &&
+        com.openminis.app.data.FastModePrefs.isEnabled() &&
+        model.id.contains("gpt", ignoreCase = true)
+    ) {
+        body.put("service_tier", "priority")
+    }
 
     if (systemPrompt != null) {
         body.put("instructions", systemPrompt)

src/android/app/src/test/java/com/openminis/app/provider/CodexFastModeTest.kt

@@ -0,0 +1,84 @@
+package com.openminis.app.provider
+
+import com.openminis.app.data.FastModePrefs
+import com.openminis.app.data.model.LLMMessage
+import com.openminis.app.data.model.LLMModel
+import com.openminis.app.provider.openai.OpenAIProvider
+import com.openminis.app.provider.openai.buildRequestBody
+import com.openminis.app.provider.openai.buildRequestBody
+import com.openminis.app.provider.openai.buildResponsesAPIBody
+import org.json.JSONObject
+import org.junit.After
+import org.junit.Assert.assertEquals
+import org.junit.Assert.assertFalse
+import org.junit.Test
+
+/**
+ * [T-codex-fast-mode] Fast Mode on the Responses path for GPT models.
+ *
+ * The Responses builder sends `service_tier: "priority"` when the app-level toggle is on and
+ * the model is a gpt-family model (OpenMinis 1.14 behaviour). It had been replaced by the
+ * xAI per-instance call alone, so the toggle did nothing for GPT. xAI's own gate is covered
+ * by [XAIPriorityProcessingTest].
+ */
+class CodexFastModeTest {
+
+    @After
+    fun tearDown() {
+        FastModePrefs.setCachedEnabledForTest(false)
+    }
+
+    private val history = listOf(LLMMessage(LLMMessage.Role.USER, "hello"))
+
+    private fun provider(modelId: String, useResponsesAPI: Boolean = true) = OpenAIProvider(
+        apiKey = "test-key",
+        model = LLMModel(id = modelId, displayName = modelId, provider = "OpenAI"),
+        basePath = "https://example.invalid/v1",
+        useResponsesAPI = useResponsesAPI,
+    )
+
+    private fun responsesBody(modelId: String, fastMode: Boolean): JSONObject {
+        FastModePrefs.setCachedEnabledForTest(fastMode)
+        return provider(modelId).buildResponsesAPIBody(
+            messages = history,
+            systemPrompt = null,
+            maxTokens = 1024,
+            stream = false,
+        )
+    }
+
+    @Test
+    fun `a gpt model on the Responses path carries priority when Fast Mode is on`() {
+        assertEquals("priority", responsesBody("gpt-6-luna", fastMode = true).getString("service_tier"))
+    }
+
+    @Test
+    fun `the tier is absent when Fast Mode is off`() {
+        assertFalse(responsesBody("gpt-6-luna", fastMode = false).has("service_tier"))
+    }
+
+    @Test
+    fun `a model outside the gpt family never carries the tier`() {
+        assertFalse(
+            "service_tier leaked into a non-GPT model's Responses body",
+            responsesBody("deepseek-v4-pro", fastMode = true).has("service_tier"),
+        )
+    }
+
+    @Test
+    fun `Chat Completions for a gpt model does not carry the tier`() {
+        FastModePrefs.setCachedEnabledForTest(true)
+        val body = provider("gpt-6-luna", useResponsesAPI = false).buildRequestBody(
+            messages = history,
+            systemPrompt = null,
+            maxTokens = 1024,
+            stream = false,
+            temperature = null,
+            imageParts = emptyList(),
+        )
+        assertFalse(
+            "the Chat Completions body only carries the tier for an eligible (xAI) instance",
+            body.has("service_tier"),
+        )
+    }
+}